Idiots Guide to Running Fixed Effects Regression

[Chthoniid]

Hi

Not sure why the panel and pooled regressions don't seem intuitive to me in Eviews. After struggling to create workfile (and a big thanks to the FAQ here on importing excel for panel/pooled data), I just can't seem to run a basic Fixed Effects model.

As soon as I had the dummy variables for the (four) cross-sectional entities, it crashes and burns. Rather it returns a near-singular matrix error message. With even just one cross-sectional entity. I know this probably sounds idiotic, but how are we supposed to incorporate the fixed-effects dummy variables into the specification?
TIA B

[EViews Gareth]

Don't include dummies, but use the fixed effects option when estimating.

[Chthoniid]

Then how are we supposed to estimate the value of the coefficients for the cross-sectional dummy variables and test them? I don't get the intuition here.

There are (in file attached) 4 cross-sectional entities (iron, rubber, stone, textile). If I exclude them, the coefficients for employ and overtime don't match the textbook (Halcoussis) answers (Table 9-G).

[EViews Gareth]

You can view the values of the fixed effects from the View menu.

If you want to perform tests on them, you'll have to manually enter dummy variables. But be careful of the dummy variable trap.

[Chthoniid]

Sorry- worked it out. The cross section stuff appears in the "View" menu and not the "Stats" output menu